运用主成分变换-逆变换对ADEOS AVNIRXS和PAN进行复合的研究

信息复合的目的是为了把多种遥感器数据的特性进行互补。复合图像与原图像相比,既具有多光谱特性,而且空间分辨率也得到了提高,这样它更有利于目视判读或计算机自动化处理。运用主成分变换-逆变换的方法对ADEOSAVNIR的多光谱数据(16 m)和全色波段数据(8 m)进行了复合研究。即先对多光谱图像进行主成分变换,求出第一主成分。然后用全色波段数据取代第一主成分并进行主成分逆变换得到所需的复合图像。复合图像的空间分辨率为8 m,而且具有4个波段。

Fusion of ADEOS-AVNIR Panchromatic and Multispectral Image Data Using Principal Component Analysis

By fusing ADEOS AVNIR panchromatic channel(PAN) with 8m resolution into its multispectral channel(XS) with 16m resolution,it is possible to produce a high resolution image suitable for photo integration.The principal component analysis(PCA) method has been tested in order to produce color composite image with a high spatial resolution using ADEOS AVNIR's multispectral mode(16m,four channels) in combination with its 8m panchromatic mode.The fusion process involves two steps:1)Geometric calibration of the AVNIR XS data and PAN data.the geometric error is less than one pixel size(<8m);2)Conducting principal component analysis and reverse principal component analysis to merge the XS and PAN data.Both visual interpretation and statistical index(i.e.Min-max,mean,median,standard deviation and information quantity) are applied to analyze the fusion effect.The results show that the rich multispectral content are maintained after fusion while the spatial resolution increased.